Two major therapeutic challenges in alcohol dependence are to minimize neuronal damage that may occur during detoxification, and to prevent relapse after detoxification. The project uses a combination of basic science and drug discovery approaches to address these challenges, with the primary aim being to develop a hierarchy of screens and models that will facilitate the identification and investigation of drugs of potential use against these aspects of dependence. Since glutamate/NMDA receptors (NMDARs) are implicated in both, via central roles in neuronal excitotoxicity and in conditioned cue-induced relapse, we will use compounds that inhibit NMDAR function in developing the screens. The therapeutic value of most NMDAR antagonists is limited by abuse potential and side-effects, but drugs that modulate the NMDAR via glycine or polyamine sites are believed not to share these characteristics. In addition to known or putative ligands for these modulatory sites (e.g. cyclo-serine, aminocyclopropanecarboxylic acid, ifenprodil, agmatine and acamprosate), we are generating libraries of synthetic and natural compounds that are structurally related to these. From these libraries we will use a high throughput pharmacological screen for radioligand binding to the glycine and polyamine sites to identify compounds that modulate NMDARs. These compounds will then be compared with the known modulators in cellular and behavioral tests based on the working hypothesis that inhibitory modulation of the NMDAR will reduce: (i) excess neurotoxicity associated with alcohol withdrawal in vitro, (ii) immediate and delayed consequences of alcohol withdrawal in animals in vivo, (iii) animal behavior elicited by cues associated with repeated alcohol administration, and (iv) psychophysiological responses in humans elicited by alcohol-related cues. These experiments will establish a hierarchy of tests of increasing complexity to evaluate compounds of potential therapeutic use. Although, it is not yet possible to establish predictive therapeutic validity for the screens, it is possible to establish that the very simple screens have predictive validity in progressively more complex models (including human models), and that these in turn have face validity for the therapeutic situation. Regardless of whether novel therapeutic compounds are discovered, the project will lay the foundations for a screening approach to medications development for alcohol dependence. The database that we develop will also test the hypothesis that NMDARs are involved in alcohol withdrawal toxicity and conditioned behavior.